Engineered cells have the potential to serve as powerful smart therapeutic agents, but we lack a systematic framework for engineering cells to optimize their efficacy and minimize their adverse effects. For example, T cells expressing artificial tumor-directed receptors (Chimeric Antigen Receptors - CARs) have achieved unexpected success in treating certain leukemias. But the engineered cells lack the precision and control of native T cells: off-target cell killing as well as hyperactivation of the immune system lead to severe adverse effects in most patients. Achieving control over these engineered T cells remains a major challenge to making this a safe and reliable therapy that can be extended to a wider range of diseases. We propose to apply the approaches of synthetic biology to systematically engineer therapeutic immune cells with far more tailored and precisely controlled behaviors. We will apply our knowledge of cell signaling to engineer modular switches and circuits that can be used to control these therapeutic cells. Our project will span two complementary approaches, represented by the following aims: Aim 1. Develop toolbox of circuit modules for precision control of CAR T cells. We will construct modules that can be combined with CARs to increase their specificity and safety, including: a. Remote control of T cells via small molecules or light. Provide physicians with safety switches for controlling activiy of transplanted therapeutic cells. b. Tunable feedback circuits. Optimize CAR T cell responses by tuning dose-response thresholds, kinetic proofreading, and the amplitude/duration of response. c. Multi-antigen triggered activation. Increase discrimination of cancer cell targeting. d. User-guided cell migration. Target T cells to specific sites of action, increasing precision. e. Synthetic cell-cell communication. Engineer autocrine and multi-cell decision-making circuits. Aim 2. Refactoring T cells: identify submodules in the T cell response network that can be used to construct custom response programs. Current CARs replicate the native program of T cell activation, but in response to a novel input. We postulate that there are many alternative ways to plug into the T cell signaling network and to generate custom tailored responses. We will scan combinatorial libraries of rewired T cell signaling proteins to identify alternative control points that can be activated through induced complex assembly (light or drugs). This network-rewiring library should reveal modular subnetworks that can be used as building blocks to construct custom response programs, including: a. Separate inducible control of: i) T cell proliferation, ii) tumor cell killing, and iii) memory cell establishment b. Induction of alternatie cytokine response programs ideal for cancer treatment (e.g. induce Th1 program) c. Induction of custom a la carte hybrid cytokine response programs that are optimized for cancer treatment.